Re: [Xen-devel] [PATCH 2/3] x86 / hvm: add domain_relinquish_resources() method

["Durrant, Paul" <pdurrant@xxxxxxxxxxxx>]

> -----Original Message-----
> From: Jan Beulich <jbeulich@xxxxxxxx>
> Sent: 22 January 2020 16:01
> To: Durrant, Paul <pdurrant@xxxxxxxxxxxx>
> Cc: xen-devel@xxxxxxxxxxxxxxxxxxxx; Andrew Cooper
> <andrew.cooper3@xxxxxxxxxx>; Wei Liu <wl@xxxxxxx>; Roger Pau Monné
> <roger.pau@xxxxxxxxxx>; Jun Nakajima <jun.nakajima@xxxxxxxxx>; Kevin Tian
> <kevin.tian@xxxxxxxxx>
> Subject: Re: [PATCH 2/3] x86 / hvm: add domain_relinquish_resources()
> method
> 
> On 22.01.2020 16:56, Durrant, Paul wrote:
> >> -----Original Message-----
> >> From: Jan Beulich <jbeulich@xxxxxxxx>
> >> Sent: 22 January 2020 15:51
> >> To: Durrant, Paul <pdurrant@xxxxxxxxxxxx>
> >> Cc: xen-devel@xxxxxxxxxxxxxxxxxxxx; Andrew Cooper
> >> <andrew.cooper3@xxxxxxxxxx>; Wei Liu <wl@xxxxxxx>; Roger Pau Monné
> >> <roger.pau@xxxxxxxxxx>; Jun Nakajima <jun.nakajima@xxxxxxxxx>; Kevin
> Tian
> >> <kevin.tian@xxxxxxxxx>
> >> Subject: Re: [PATCH 2/3] x86 / hvm: add domain_relinquish_resources()
> >> method
> >>
> >> On 21.01.2020 13:00, Paul Durrant wrote:
> >>> There are two functions in hvm.c to deal with tear-down and a domain:
> >>> hvm_domain_relinquish_resources() and hvm_domain_destroy(). However,
> >> only
> >>> the latter has an associated method in 'hvm_funcs'. This patch adds
> >>> a method for the former and stub definitions for SVM and VMX.
> >>
> >> Why the stubs? Simply ...
> >>
> >>> --- a/xen/arch/x86/hvm/hvm.c
> >>> +++ b/xen/arch/x86/hvm/hvm.c
> >>> @@ -715,6 +715,8 @@ int hvm_domain_initialise(struct domain *d)
> >>>
> >>>  void hvm_domain_relinquish_resources(struct domain *d)
> >>>  {
> >>> +    hvm_funcs.domain_relinquish_resources(d);
> >>
> >> ... stick a NULL check around this one. I also wonder whether, it
> >> being entirely new, this wouldn't better use alternative call
> >> patching right from the beginning. It's not the hottest path, but
> >> avoiding indirect calls seems quite desirable, especially when
> >> doing so is relatively cheap.
> >>
> >
> > I'd like it to align with the rest of the hvm_funcs so I'll add the
> > NULL check, but alternatives patch for all hvm_funcs seems like a
> > good thing I the longer term.
> 
> The frequently used ones have been converted already. Hence my
> suggestion to make new ones use that model from the beginning.
> 

Oh, ok. I'll go look for some examples.

  Paul
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel